Structure of an invertebrate gene encoding cytoplasmic intermediate filament (IF) proteins: implications for the origin and the diversification of IF proteins.

The structure of the single gene encoding the cytoplasmic intermediate filament (IF) proteins in non-neuronal cells of the gastropod Helix aspersa is described. Genomic and cDNA sequences show that the gene is composed of 10 introns and 11 exons, spanning greater than 60 kb of DNA. Alternative RNA processing accounts for two mRNA families which encode two IF proteins differing only in their C-terminal sequence. The intron/exon organization of the Helix rod domain is identical to that of the vertebrate type III IF genes in spite of low overall protein sequence homology and the presence of an additional 42 residues in coil 1b of the invertebrate sequence. Intron position homology extends to the entire coding sequence comprising both the rod and tail domains when the invertebrate IF gene is compared with the nuclear lamin LIII gene of Xenopus laevis presented in the accompanying report of Döring and Stick. In contrast the intron patterns of the tail domains of the invertebrate IF and the lamin genes differ from those of the vertebrate type III genes. The combined data are in line with an evolutionary descent of cytoplasmic IF proteins from a nuclear lamin-like progenitor and suggest a mechanism for this derivation. The unique position of intron 7 in the Helix IF gene indicates that the archetype IF gene arose by the elimination of the nuclear localization sequence due to the recruitment of a novel splice site. The presumptive structural organization of the archetype IF gene allows predictions with respect to the later diversification of metazoan IF genes. Whereas models proposing a direct derivation of neurofilament genes seem unlikely, the earlier speculation of an mRNA transposition mechanism is compatible with current results.     

Analysis of the cDNA and gene encoding a cytoplasmic intermediate filament (IF) protein from the cephalochordate Branchiostoma lanceolatum; implications for the evolution of the IF protein family.

We report the molecular cloning of a full-length cDNA encoding a non-neuronal cytoplasmic intermediate filament (IF) protein of the cephalochordate Branchiostoma lanceolatum. Sequence and structural characteristics of IF-1 reveal a close relation to vertebrate IF proteins: they all lack the extended coil 1b version and the lamin tail homology found in protostomic IF proteins. This implies that divergence of type I to IV IF genes from a common ancestor either coincided with the origin of chordates or occurred at an earlier stage in the evolution of deuterostomes. The structural organization of the cephalochordate gene shows a closer relation to vertebrate type III genes than to type I or II genes. The single gene (approximately 19 kb) is composed of 7 exons and 6 introns which are all located within the sequence encoding the rod domain. The positions and phases of the introns show perfect homology to vertebrate type III genes. In line with the absence of protein sequence similarity of the tail domain, the Branchiostoma gene does not possess the introns interrupting this region in type III genes of vertebrates.     

Gene structure of nuclear lamin LIII of Xenopus laevis; a model for the evolution of IF proteins from a lamin-like ancestor.

The lamin LIII gene of Xenopus laevis has been characterized. The gene is duplicated in the Xenopus genome. The transcribed region spreads over 22 kb of genomic DNA encoding 12 exons. Two alternatively spliced mRNAs are observed which encode LIII isoforms that differ only by the 12 C-terminal amino acids which, however, both contain the CaaX motif known to be the target of post-translational modifications. The intron pattern of the lamin LIII gene is strikingly similar to that of an invertebrate intermediate filament (IF) gene over the entire protein coding sequence. The similarity in gene structure is restricted to the rod domain when compared with vertebrate types I-III IF genes. Our data suggest a model of how IF proteins evolved from a lamin-like ancestor by deletion of two signal sequences; the nuclear localization signal and the C-terminal ras-related CaaX motif. The data rule out the previously proposed hypothesis that IF proteins evolved from an intronless ancestor with an early divergence of neuronal and non-neuronal IF proteins. Together with the data presented in the accompanying paper by Dodemond et al. it can be concluded that the tail domains of lamins and invertebrate IF proteins, but not those of vertebrate IF proteins, are homologous. Thus, the different vertebrate IF proteins probably evolved by combination of the central rod domain with different tail domains by exon shuffling.     

Sequence and structure of the mouse gene coding for the largest neurofilament subunit

We have determined the complete nucleotide sequence of the mouse gene encoding the neurofilament NF-H protein. The C-terminal domain of NF-H is very rich in charged amino acids (aa) and contains a 3-aa sequence, Lys-Ser-Pro, that is repeated 51 times within a stretch of 368 aa. The location of this serine-rich repeat in the phosphorylated domain of NF-H indicates that it represents the major protein kinase recognition site. The nfh gene shares two common intron positions with the nfl and nfm genes, but has an additional intron that occurs at a location equivalent to one of the introns in non-neuronal intermediate filament-coding genes. This additional nfh intron may have been acquired via duplication of a primordial intermediate filament gene.     

The gene for a cytoplasmic intermediate filament (IF) protein of the hemichordate Saccoglossus kowalevskii; definition of the unique features of chordate IF proteins

We have isolated full length cDNAs encoding a cytoplasmic intermediate filament (IF) protein of a priapulid (Priapulus caudatus) and of a hemichordate (Saccoglossus kowalevskii) and determined the organisation of the hemichordate gene. As expected, the priapulid protein shows the long coil 1b subdomain and the lamin tail homology segment already known for cytoplasmic IF proteins from 11 other protostomic phyla. Surprisingly, the hemichordate protein follows in domain organisation much more closely the protostomic IF proteins than the chordate IF proteins. Thus, the lack of a lamin tail homology segment together with a deletion of 42 residues in the coil 1b domain is a molecular feature restricted to the chordates. We propose that the known IF subfamilies I to IV developed by gene duplications and sequence drift after the deletion in coil 1b arose at the base of the chordate branch.     

Structure and evolutionary origin of the gene encoding mouse NF-M, the middle-molecular-mass neurofilament protein

We describe the complete sequence of the gene encoding mouse NF-M, the middle-molecular-mass neurofilament protein. The coding sequence is interrupted by two intervening sequences which align perfectly with the first two intervening sequences in the gene encoding NF-L (the low-molecular-mass neurofilament protein); there is no intron in the gene encoding NF-M corresponding to the third intron in NF-L. Therefore, both the number of introns and their arrangement in the genes coding NF-L and NF-M contrast sharply with the number and arrangement of introns in the genes of known sequence, encoding other members of the intermediate filament multigene family (desmin, vimentin, glial fibrillary acidic protein and the acidic and basic keratins); with the exception of a single truncated keratin gene that lacks an encoded tailpiece, these genes all contain eight introns, of which at least six are placed at homologous locations. Assuming the existence of a primordial intermediate filament gene containing most (if not all) the introns found in contemporary non-neurofilament intermediate filament genes, it seems likely that an RNA-mediated transposition event was involved in the generation of an ancestral gene encoding the NF polypeptides. A combination of insertional transposition and gene-duplication events could then explain the anomalous number and placement of introns within these genes. Consistent with this notion, we show that the genes encoding NF-M and NF-L are linked.     

An amplified insect dihydrofolate reductase gene contains a single intron

We have used methotrexate-resistant mosquito (Aedes albopictus) cells as the source of DNA for cloning an 8.5-kb EcoRI fragment containing an amplified dihydrofolate reductase (DHRF) gene. An estimated 1200 copies of the DHFR gene were represented in nuclear DNA from Mtx-5011-256 cells, which were 3000-fold more resistant to methotrexate than wild-type cells. Southern blot analysis indicated that all of the amplified DHFR genes were contained within a 1.8-kb AccI fragment represented in the cloned DNA. In contrast to mammalian DHFR genes which span approximately 30 kb, the complete amino acid coding sequence of the mosquito DHFR gene spanned 614 nucleotides, including a single 56-nucleotide intron that interrupted a conserved Arg codon at amino acid position 27. Additional introns characteristic of mammalian DHFR genes were absent; conservation of the first intron in the mosquito DHFR gene supports a regulatory role for this intron. The mosquito DHFR gene coded for a 186-amino-acid protein with 43-48% similarity to vertebrate DHFR.     

Lamins of the sea lamprey (Petromyzon marinus) and the evolution of the vertebrate lamin protein family

Lamin proteins are found in all metazoans. Most non-vertebrate genomes including those of the closest relatives of vertebrates, the cephalochordates and tunicates, encode only a single lamin. In teleosts and tetrapods the number of lamin genes has quadrupled. They can be divided into four sub-types, lmnb1, lmnb2, LIII, and lmna, each characterized by particular features and functional differentiations. Little is known when during vertebrate evolution these features have emerged. Lampreys belong to the Agnatha, the sister group of the Gnathostomata. They split off first within the vertebrate lineage. Analysis of the sea lamprey (Petromyzon marinus) lamin complement presented here, identified three functional lamin genes, one encoding a lamin LIII, indicating that the characteristic gene structure of this subtype had been established prior to the agnathan/gnathostome split. Two other genes encode lamins for which orthology to gnathostome lamins cannot be designated. Search for lamin gene sequences in all vertebrate taxa for which sufficient sequence data are available reveals the evolutionary time frame in which specific features of the vertebrate lamins were established. Structural features characteristic for A-type lamins are not found in the lamprey genome. In contrast, lmna genes are present in all gnathostome lineages suggesting that this gene evolved with the emergence of the gnathostomes. The analysis of lamin gene neighborhoods reveals noticeable similarities between the different vertebrate lamin genes supporting the hypothesis that they emerged due to two rounds of whole genome duplication and makes clear that an orthologous relationship between a particular vertebrate paralog and lamins outside the vertebrate lineage cannot be established.     

Crystal structure of the human lamin A coil 2B dimer: implications for the head-to-tail association of nuclear lamins

Nuclear intermediate filaments (IFs) are made from fibrous proteins termed lamins that assemble, in association with several transmembrane proteins of the inner nuclear membrane and an unknown number of chromatin proteins, into a filamentous scaffold called the nuclear lamina. In man, three types of lamins with significant sequence identity, i.e. lamin A/C, lamin B1 and B2, are expressed. The molecular characteristics of the filaments they form and the details of the assembly mechanism are still largely unknown. Here we report the crystal structure of the coiled-coil dimer from the second half of coil 2 from human lamin A at 2.2A resolution. Comparison to the recently solved structure of the homologous segment of human vimentin reveals a similar overall structure but a different distribution of charged residues and a different pattern of intra- and interhelical salt bridges. These features may explain, at least in part, the differences observed between the lamin and vimentin assembly pathways. Employing a modeled lamin A coil 1A dimer, we propose that the head-to-tail association of two lamin dimers involves strong electrostatic attractions of distinct clusters of negative charge located on the opposite ends of the rod domain with arginine clusters in the head domain and the first segment of the tail domain. Moreover, lamin A mutations, including several in coil 2B, have been associated with human laminopathies. Based on our data most of these mutations are unlikely to alter the structure of the dimer but may affect essential molecular interactions occurring in later stages of filament assembly and lamina formation.     

Complete sequence of a bovine type I cytokeratin gene: conserved and variable intron positions in genes of polypeptides of the same cytokeratin subfamily

The complete sequence of a bovine gene encoding an epidermal cytokeratin of mol. wt. 54 500 (No VIb) of the acidic (type I) subfamily is presented, including an extended 5' upstream region. The gene (4377 bp, seven introns) which codes for a representative of the glycine-rich subtype of cytokeratins of this subfamily, is compared with genes coding for: another subtype of type I cytokeratin; a basic (type II) cytokeratin gene; and vimentin, a representative of another intermediate filament (IF) protein class. The positions of the five introns located within the highly homologous alpha-helix-rich rod domain are identical or equivalent, i.e., within the same triplet, in the two cytokeratin I genes. Four of these intron positions are also identical with intron sites in the vimentin gene, and three of these intron positions are identical or similar in the type I and type II cytokeratin subfamilies. On the other hand, the gene organization of both type I cytokeratins differs from that of the type II cytokeratin in the rod region in five intron positions and in the introns located in the carboxy-terminal tail region, with the exception of one position at the rod-tail junction. Remarkably, the two type I cytokeratins also differ from each other in the positions of two introns located at and in the region coding for the hypervariable, carboxy-terminal portion. The introns and the 5' upstream regions of the cytokeratin VIb gene do not display notable sequence homologies with the other IF protein genes, but sequences identical with--or very similar to--certain viral and immunoglobulin enhancers have been identified.(ABSTRACT TRUNCATED AT 250 WORDS)     

Characterization of lamin mutation phenotypes in Drosophila and comparison to human laminopathies

Lamins are intermediate filament proteins that make up the nuclear lamina, a matrix underlying the nuclear membrane in all metazoan cells that is important for nuclear form and function. Vertebrate A-type lamins are expressed in differentiating cells, while B-type lamins are expressed ubiquitously. Drosophila has two lamin genes that are expressed in A- and B-type patterns, and it is assumed that similarly expressed lamins perform similar functions. However, Drosophila and vertebrate lamins are not orthologous, and their expression patterns evolved independently. It is therefore of interest to examine the effects of mutations in lamin genes. Mutations in the mammalian lamin A/C gene cause a range of diseases, collectively called laminopathies, that include muscular dystrophies and premature aging disorders. We compared the sequences of lamin genes from different species, and we have characterized larval and adult phenotypes in Drosophila bearing mutations in the lam gene that is expressed in the B-type pattern. Larvae move less and show subtle muscle defects, and surviving lam adults are flightless and walk like aged wild-type flies, suggesting that lam phenotypes might result from neuromuscular defects, premature aging, or both. The resemblance of Drosophila lam phenotypes to human laminopathies suggests that some lamin functions may be performed by differently expressed genes in flies and mammals. Such still-unknown functions thus would not be dependent on lamin gene expression pattern, suggesting the presence of other lamin functions that are expression dependent. Our results illustrate a complex interplay between lamin gene expression and function through evolution.     

Characterization of the Hydra Lamin and Its Gene: A Molecular Phylogeny of Metazoan Lamins

We report sequences for nuclear lamins from the teleost fish Danio and six invertebrates. These include two cnidarians (Hydra and Tealia), one priapulid, two echinoderms, and the cephalochordate Branchiostoma. Combining these results with earlier data on Drosophila, Caenorhabditis elegans, and various vertebrates, the following conclusions on lamin evolution can be drawn. First, all invertebrate lamins resemble in size the vertebrate B-type lamin. Second, all lamins described previously for amphibia, birds and mammals as well as the first lamin of a fish, characterized here, show a cluster of 7 to 12 acidic residues in the tail domain. Since this acidic cluster is absent from all invertebrate lamins including that of the cephalochordate Branchiostoma, it was acquired with the vertebrate lineage. The larger A-type lamin of differentiated cells must have arisen subsequently by gene duplication and insertion of an extra exon. This extra exon of the vertebrate A-lamins is the only major change in domain organization in metazoan lamin evolution. Third, the three introns of the Hydra and Priapulus genes correspond in position to the last three introns of vertebrate B-type lamin genes. Thus the entirely different gene organization of the C. elegans and Drosophila Dmo genes seems to reflect evolutionary drift, which probably also accounts for the fact that C. elegans has the most diverse lamin sequence. Finally we discuss the possibility that two lamin types, a constitutively expressed one and a developmentally regulated one, arose independently on the arthropod and vertebrate lineages. Received: 4 February 1999 / Accepted: 1 April 1999     

The single nuclear lamin of Caenorhabditis elegans forms in vitro stable intermediate filaments and paracrystals with a reduced axial periodicity

The lamins of the tunicate Ciona intestinalis and the nematode Caenorhabditis elegans show unusual sequence features when compared to the more than 35 metazoan lamin sequences currently known. We therefore analyzed the in vitro assembly of these two lamins by electron microscopy using chicken lamin B2 as a control. While lamin dimers usually appear as a rod carrying two globules at one end, these globules are absent from Ciona lamin, which lacks the central 105-residue region of the tail domain. The deletion of 14 residues or two heptads from the coiled coil rod domain of the single C.elegans lamin results in a 1.5-nm shortening of the dimer rod. Similarly, the paracrystals assembled from the C.elegans lamin exhibit a 3.1-nm reduction of the true axial repeat compared to that of chicken lamin B2 paracrystals. We speculate that the banding pattern in the C.elegans lamin paracrystals arises from a relative stagger between dimers and/or a positioning of the globular tail domain relative to the central rod that is distinct from that observed in chicken lamin B2 paracrystals. Here we show that a nuclear lamin can assemble in vitro into 10-nm intermediate filaments (IFs). C.elegans lamin in low ionic strength Tris-buffers at a pH of 7.2-7.4 provides a stable population of lamin IFs. Some implications of this filament formation are discussed.     

In contrast to the nematode and fruit fly all 9 intron positions of the sea anemone lamin gene are conserved in human lamin genes

We identified the single gene for nuclear lamin in the genome draft of the sea anemone Nematostella vectensis, a member of the cnidaria, a very old metazoan phylum. The gene consists of 10 exons and 9 introns. Strikingly all 9 intron positions are conserved in the human lamin B genes, which have only 1 (lamin B1) or 2 (lamin B2) additional introns. Using the information on neighboring genes we propose that the human lamin B1 gene on chromosome 5 is the true homolog of the Nematostella lamin gene, while the lamin B2 gene on chromosome 19 arose during vertebrate evolution. In marked contrast to this conservation of gene structure are the results in the rapidly evolving genomes of Drosophila and Caenorhabditis elegans. Here the lamin genes have much fewer introns and these occur often at novel positions. In the single nematode lamin gene and the Drosophila lamin Dmo gene no intron position coincides with an intron in the sea anemone lamin gene.     

Rapidly evolving lamins in a chordate, Oikopleura dioica, with unusual nuclear architecture

Metazoan lamins are implicated in the organization of numerous critical nuclear processes. Among chordates, the appendicularian, Oikopleura dioica, has an unusually short life cycle involving rapid growth through extensive recourse to endoreduplication, a characteristic more associated with some invertebrates. In some tissues, this is accompanied by the formation of elaborate, bilaterally symmetric nuclear morphologies associated with specific gene expression patterns. Lamin composition can mediate nuclear shape in spermiogenesis as well as during pathological and normal aging and we have analyzed the O. dioica lamin and intermediate filament (IF) complement, comparing it to that present in other deuterostomes. O. dioica has one lamin gene coding two splice variants. Both variants share with the sister class ascidians a highly reduced C-terminal tail region lacking the immunoglobulin fold, indicating this derivation occurred at the base of the urochordate lineage. The OdLamin2 variant has a unique insertion of 63 amino acids in the normally short N-terminal region and has a developmental expression profile corresponding to the occurrence of endocycling. O. dioica has 4 cytoplasmic IF proteins, IF-A, C, Dalpha, and Dbeta. No homologues to the ascidian IF-B or F proteins were identified, reinforcing the suggestion that these proteins are unique to ascidians. The degree of sequence evolution in the rod domains of O. dioica cytoplasmic IF proteins and their closest ascidian and vertebrate homologues was similar. In contrast, whereas the rate of lamin B rod domain sequence evolution has also been similar in vertebrates, cephalochordates and the sea urchin, faster rates have occurred among the urochordates, with the O. dioica lamin displaying a far greater rate than any other lamin.